Light

soumith / galaxyzoo Goto Github PK

View Code? Open in Web Editor NEW

35.0 6.0 12.0 447 KB

Entry for GalaxyZoo challenge

License: Other

Lua 98.50% Shell 1.50%

galaxyzoo's Introduction

galaxyzoo

Entry for GalaxyZoo Challenge: http://www.kaggle.com/c/galaxy-zoo-the-galaxy-challenge

Our best score was 0.08246 placing us at 9th position, we trained the network for about 6 days, 4 models with different random seeds were trained on 2 NVidia Titan cards. The model that was trained for entry is in 2_model.lua

The winning model was Sander Dieleman's convnet appropach described here: http://benanne.github.io/2014/04/05/galaxy-zoo.html

For install instructions, please look at INSTALL.md

galaxyzoo's People

Contributors

Stargazers

Watchers

Forkers

ageek chenglongchen fish444555 jakobularius lixunlove rtvt123 zhmz90 davidkh1 andrubrown vpm238 kwccoin

galaxyzoo's Issues

decide on adjustment of all the knobs (learning rate etc.)

Install Torch on CentOS without sudo

Hi Soumith,

What is the best way to install Torch on Cent OS 6.7 without sudo rights? Thanks!

Cheers,
Zhong

make it GPU ready. CPU wont cut it in terms of training time for the deadline.

input mean subtraction

create a data sampler

make the testing on a wide variety of rotations as well (for each test sample)

train an overfeat-based model as well, for fun

load data using the csv package

create the training script

create the decision tree graph

write the decoder script that outputs things in the format required by kaggle

change the cost function to the unnormalized vector

make dropout optional

verify all-zeros and all-ones score

make the training have individual MSE criterion for each question

decide on the feature architecture

add artificial jitter modules

Which CSV package are you using?

The one installed I from Luarocks only has the "open" and "use" function, rather than the load function you call in 1_data.lua file.

howto accumulate error when `cirterion:cuda()` is used

hi, ~
currently, I'm learning torch7 according to this repo, especially for implementation on GPU. And I'm confused how to reduce all errors calcualted by each kernel after cirtierion:cuda() is called. In this repo, this usage is commented.
Can anyone give me some advice or hints about criterion:cuda() ? thanks~

write a converter that normalizes the probabilities of each question for training, and write a reverse mapper as well

start training with big network, not much time left for deadline :-/

Recommend Projects

React

A declarative, efficient, and flexible JavaScript library for building user interfaces.
Vue.js

🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript

TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
TensorFlow

An Open Source Machine Learning Framework for Everyone
Django

The Web framework for perfectionists with deadlines.
Laravel

A PHP framework for web artisans
D3

Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

javascript

JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
web

Some thing interesting about web. New door for the world.
server

A server is a program made to process requests and deliver data to clients.
Machine learning

Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Visualization

Some thing interesting about visualization, use data art
Game

Some thing interesting about game, make everyone happy.

Recommend Org

Facebook

We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft

Open source projects and samples from Microsoft.
Google

Google ❤️ Open Source for everyone.
Alibaba

Alibaba Open Source for everyone
D3

Data-Driven Documents codes.
Tencent

China tencent open source team.